A Predictive System for Efficient Portfolio Management: An Application of ANN and Technical Indicators

Portfolio management amplifies profit and curtails risk by allocating the assets effectively. In financial market the aim of any investor is to achieve high profit. It is very important for an investor to buy and sell their investment at appropriate time to maximize their profit. But the nature of the stock market is highly fluctuating and dynamic making it hard to decide the proper time to sell and buy. With the origin of artificial intelligence, computational programs of predicting the returns have proved to be quite effective in the area of portfolio management. In this proposed work, an expert system for portfolio is developed by using five different technical indicators, artificial neural network, and an advantage of reshuffling/churning process. This work possesses five steps (1) selection of stocks through five criteria, (2) stocks filtering by technical indicators, (3) construction of portfolio of equity by artificial neural network (4) weight implementation (5) Churning/reshuffling. Ten companies of different sectors are selected in our portfolio in order to gain maximum profit. Finally, the output is compared with five star rated mutual funds. It was found that the maximum output given by top star mutual funds is 16.79% while the output of proposed system is 21.95% which is far better in the same time frame.